Sustainable bridge design by metamodel-assisted multi-objective optimization and decision-making under uncertainty

Abstract

[EN] Today, bridge design seeks not only to minimize cost, but also to minimize adverse environmental and social impacts. This multi-criteria decision-making problem is subject to variability of the opinions of stakeholders regarding the importance of criteria for sustainability. As a result, this paper proposes a method for designing and selecting optimally sustainable bridges under the uncertainty of criteria comparison. A Pareto set of solutions is obtained using a metamodel-assisted multi-objective optimization. A new decision-making technique introduces the uncertainty of the decision-maker's preference through triangular distributions and thereby ranks the sustainable bridge designs. The method is illustrated by a case study of a three-span post-tensioned concrete box-girder bridge designed according to the embodied energy, overall safety and corrosion initiation time. In this particular case, 211 efficient solutions are reduced to two preferred solutions which have a probability of being selected of 81.6% and 18.4%. In addition, a sensitivity analysis validates the influence of the uncertainty regarding the decision making. The approach proposed allows actors involved in the bridge design and decision-making to determine the best sustainable design by finding the probability of a given design being chosen.